Supporting Information
Vellend et al. 10.1073/pnas.1312779110
SI Methods
Papers, Studies, and Data Sets. Data meeting our criteria were
found in 148 papers. Six papers reported data separately for two
to three distinctly different habitat types, which we considered as
separate studies, for a total of 168 studies. Individual studies often
reported data separately for multiple sites within a single habitat
type, which we considered separate data sets, for a total of 346
data sets. All authors contributed to an initial round of data
extraction, after which each line of data were rechecked by two
authors (M.V. and L.B.), working together closely to resolve any
discrepancies or ambiguities and to ensure a standardized pro-
tocol across all papers.
Candidate Drivers of Vegetation Change. In cases for which the
authors of a paper identified a primary driver of temporal veg-
etation change (262/346 data sets), we classified drivers as follows:
ongoing disturbance, postdisturbance succession (disturbances
include severe storms, volcanos, logging, drought, disease, and
fire, with fire analyzed separately as the only subcategory with
n ≥ 5), ongoing or increased grazing/herbivory, reduced grazing/
herbivory, climate change, pollution (mostly atmospheric nutri-
ent deposition), plant invasion, increased canopy density (i.e.,
reduced light), altered hydrology, or management (this category
includes altered regimes of multiple factors simultaneously, such
as fertilization, mowing, grazing, or coppicing, when no one such
activity was identified as the primary driver). Ongoing dis-
turbances are those that occurred (often repeatedly) during the
study period, whereas postdisturbance succession indicates
studies for which the first vegetation survey was conducted after
the last disturbance event. Three individual studies had unique
drivers (urbanization, fragmentation, and seabird colonization)
and were not included in this analysis.
Statistical Analysis. We first used a multilevel approach including
Markov chain Monte Carlo (MCMC) simulations in R version
2.15.2 and OpenBUGS via the R2OpenBUGS package (1, 2).
With multiple data sets in many studies, data sets were not in-
dependent of one another, and we incorporated this dependency
into a multilevel model. For data set i in study j, the log ratio can
be rewritten as ti−1log(SRY2,i/SRY1,i) = ESj(i) + «i. The ESj are the
study-level effect sizes that come from a common distribution
ESj ∼ N(μES, τ2), in which μES is the grand mean, or population-
average trend in diversity over time. The «i term expresses the
data set level error. Thus, the analysis is a random-effect meta-
analysis, not assuming a single effect size value that is common to
all studies (3). The posterior distribution of μES, with credible
intervals expressing the uncertainty about μES, indicates whether
on average diversity is increasing or decreasing over time. The
superpopulation SD, τ, represents variation among the modeled
distribution of effect sizes from which our studies were selected,
rather than only variation within the set of studies included in
our meta-analysis.
Meta-analyses often use variance estimates to weight different
observations. However, variances around mean species richness
values were only reported in 30% of the studies in our data set,
and none reported the covariance between the nonindependent
mean SRY1,i and SRY2,i values, which is needed to calculate the
variance of the log ratio (4). Thus, we took an unweighted ap-
proach in the multilevel analysis, assuming the data-level errors
to follow a common distribution «i ∼ N(0,σ 2) (5). Minimal prior
information was introduced by using vague priors for the pa-
rameters τ and σ: both ∼U(0,100) and μES ∼ N(0,10,000) (5).
Vellend et al. www.pnas.org/cgi/content/short/1312779110
Three independent Markov chains were run for 100,000 iter-
ations, discarding the first 5,000 (burn-in), and retaining one out
of each hundred iterations (thinning) (6). This analysis was
complemented by a more conventional bootstrapped fixed-effects
meta-analysis.
A subset of studies reported diversity and/or evenness metrics;
these data were analyzed using the same multilevel approach as
for the species richness data (effect sizes). We recognize that
absolute or proportional changes in these metrics are not as easily
interpreted as changes in species richness, so these analyses fo-
cused on testing for directional trends rather than attempting
to relate any such changes to biodiversity-ecosystem function
(BDEF) experiments. In 35 of 38 studies, Shannon diversity was
reported [H = −Σpiln(pi), in which pi is the relative abundance of
species i]. For analysis, we calculated exp(H), the “numbers
equivalent” transformation of H, which is more quantitatively
comparable among studies (7). In 18 studies, the Pielou evenness
metric was presented [E = exp(H)/S, in which S is the number of
species in the plot]. We characterized the distribution of effect
sizes for each of these metrics.
Justification for Excluding Experimental Studies. Our primary goal
was to characterize temporal diversity change plant communities
that represent the real world, defined as any place that was not
manipulated experimentally for the purposes of studying vege-
tation change. As such, we deliberately did not consider exper-
imental studies, such as those that manipulate temperature, soil
nutrients, CO2, etc. Although such experiments can provide
important clues as to which species might increase or decrease in
abundance due to environmental changes under more natural
conditions, the rate of environmental change (e.g., an instant
doubling of atmospheric CO2 concentration) far exceeds the rate
of change experienced by natural communities. There is also
likely insufficient time for species that might benefit from the
modified conditions to colonize. As such, we restricted our
analysis to studies reporting temporal diversity changes under
the conditions experienced in the real world, which includes
locations across the globe experiencing a wide variety of envi-
ronmental and anthropogenic drivers of vegetation change. This
database is not a biased set of studies focusing only on pristine
vegetation (SI Text).
One type of experiment that does, in fact, mimic the real (i.e.,
nonexperimental) rate of environmental change is habitat frag-
mentation. The largest such study is the Biological Dynamics of
Forest Fragments project in Brazil, where, surprisingly, no dif-
ferences were found between intact forest and fragments of 1–100
ha in terms of plant species richness or diversity (Fisher’s α) over
22 y (8). Large trees in small fragments experienced greatly el-
evated mortality, with a consequent alteration of tree size
structure (8) and a reduction in carbon storage (9), but species
losses were offset by gains in species better adapted to disturbed
conditions. These results support both the absence of directional
species diversity trends found in our meta-analysis, and also our
contention that other effects of environmental changes on eco-
system function (in this case effects of mortality and species
turnover on carbon storage) likely render any effects of species
diversity per se negligible. A somewhat less realistic fragmenta-
tion experiment, in which patches of grassland (Kansas) of be-
tween 32 and 5,000 m2 were created, found similar results: larger
patches contained more species, but in equivalent small areas
(1-m2 plots) there was no such difference (10), although there
were compositional differences (11). These studies reveal two
1 of 6
important lessons for our study. First, the inclusion of such ex-
periments should not alter our conclusion of no net diversity
change over time. Second, the conventional wisdom that plant
1. Sturtz S, Ligges U, Gelman A (2005) R2WinBUGS: A package for running WinBUGS from
R. J Stat Softw 12(3):1–16.
2. R Core Team (2012) R: A Language and Environment for Statistical Computing (R
Foundation for Statistical Computing, Vienna).
3. Harrison F (2011) Getting started with meta-analysis. Methods Ecol Evol 2(1):1–10.
4. Lajeunessei MJ (2011) On the meta-analysis of response ratios for studies with
correlated and multi-group designs. Ecology 92(11):2049–2055.
5. Gelman A, Hill J (2007) Data Analysis Using Regression and Multilevel/Hierarchical
Models (Cambridge Univ Press, Cambridge, MA).
6. Hector A, et al. (2011) BUGS in the analysis of biodiversity experiments: Species richness
and composition are of similar importance for grassland productivity. PLoS ONE 6(3):
e17434.
SI Text
References for original studies used in the meta-analysis. Num-
bers correspond to the “Paper” column in Dataset S1.
1. Abella SR, Jaeger JF, Brewer LG (2004) Fifteen years of plant community dynamics
during a northwest Ohio oak savanna restoration. The Michigan Botanist 43(3):
117–127.
2. Adler PB, Lauenroth WK (2003) The power of time: Spatiotemporal scaling of species
diversity. Ecol Lett 6(8):749–756.
3. Altesor A, Di Landro E, May H, Ezcurra E (1998) Long-term species change in
a Uruguayan grassland. J Veg Sci 9(2):173–180.
4. Alvarez-Rogel J, Jimenez-Carceles FJ, Roca MJ, Ortiz R (2007) Changes in soils and
vegetation in a Mediterranean coastal salt marsh impacted by human activities.
Estuar Coast Shelf Sci 73(3-4):510–526.
5. Anderson PD (2008) Understory vegetation responses to initial thinning of Douglas-fir
plantations undergoing conversion to uneven-age management. US Department of
Agriculture-Forest Service General Technical Report PNW (US Department of Agri-
culture, Washington, DC), pp 77-90.
6. Armentano TV, et al. (2006) Rapid responses of vegetation to hydrological changes in
Taylor Slough, Everglades National Park, Florida, USA. Hydrobiologia 569(1):293–309.
7. Baeten L, et al. (2009) Herb layer changes (1954-2000) related to the conversion of
coppice-with-standards forest and soil acidification. Appl Veg Sci 12(2):187–197.
8. Baeten L, Hermy M, Van Daele S, Verheyen K (2010) Unexpected understorey
community development after 30 years in ancient and post-agricultural forests. J Ecol
98(6):1447–1453.
9. Baeten L, et al. (2012) Distinguishing between turnover and nestedness in the
quantification of biotic homogenization. Biodivers Conserv 21(6):1399–1409.
10. Báez S, Collins SL (2008) Shrub invasion decreases diversity and alters community
stability in northern Chihuahuan Desert plant communities. PLoS ONE 3(6):e2332.
11. Bakker C, Blair JM, Knapp AK (2003) Does resource availability, resource
heterogeneity or species turnover mediate changes in plant species richness in
grazed grasslands? Oecologia 137(3):385–391.
12. Barkham JP (1992) The effects of management on the ground flora of ancient
woodland, Brigsteer Park Wood, Cumbria, England. Biol Conserv 60(3):167–187.
13. Barkowski JW, Kolditz K, Brumsack H, Freund H (2009) The impact of tidal
inundation on salt marsh vegetation after de-embankment on Langeoog Island,
Germany—six years time series of permanent plots. J Coast Conserv 13(4):185–206.
14. Barrett MA, Stiling P, Lopez RR (2006) Long-term changes in plant communities
influenced by key deer herbivory. Nat Areas J 26(3):235–243.
15. Barry MJ, Andreas BK, De Szalay FA (2008) Long-term plant community changes in
managed fens in Ohio, USA. Aquatic Conservation-Marine and Freshwater
Ecosystems 18(4):392–407.
16. Baumberger T, et al. (2012) Plant community changes as ecological indicator of
seabird colonies’ impacts on Mediterranean Islands. Ecol Indic 15(1):76–84.
17. Bennie J, Hill MO, Baxter R, Huntley B (2006) Influence of slope and aspect on long-
term vegetation change in British chalk grasslands. J Ecol 94(2):355–368.
18. Berlin GAI, Linusson AC, Olsson EGA (2000) Vegetation changes in semi-natural
meadows with unchanged management in southern Sweden, 1965-1990. Acta Oecol
21(2):125–138.
19. Bowles M, Apfelbaum S, Haney A, Lehnhardt S, Post T (2011) Canopy cover and
groundlayer vegetation dynamics in a fire managed eastern sand savanna. For Ecol
Manage 262(11):1972–1982.
20. Bowles M, Jones M (2006) Testing the efficacy of species richness and floristic quality
assessment of quality, temporal change, and fire effects in tallgrass prairie natural
areas. Nat Areas J 26(1):17–30.
21. Bowles M, et al. (2003) Twenty-year woody vegetation changes in northern
flatwoods and mesic forest at Ryerson Conservation Area, Lake County, Illinois.
Erigenia 19:31–51.
22. Bowles M, McBride J, Stoynoff N, Johnson K (1996) Temporal changes in vege-
tation composition and structure in a fire-managed prairie fen. Nat Areas J 16(4):
275–288.
23. Bowles ML, Jones MD, McBride JL (2003) Twenty-year changes in burned and
unburned sand prairie remnants in northwestern Illinois and implications for
management. Am Midl Nat 149(1):35–45.
Vellend et al. www.pnas.org/cgi/content/short/1312779110
biodiversity generally declines in response to certain drivers of
change, such as habitat fragmentation, may not apply to small
spatial scales.
7. Jost L (2007) Partitioning diversity into independent alpha and beta components.
Ecology 88(10):2427–2439.
8. Laurance WF, et al. (2006) Rapid decay of tree-community composition in Amazonian
forest fragments. Proc Natl Acad Sci USA 103(50):19010–19014.
9. Nascimento HEM, Laurance WF (2004) Biomass dynamics in Amazonian forest
fragments. Ecol Appl 14(sp4):127–138.
10. Holt RD, Robinson GR, Gaines MS (1995) Vegetation dynamics in an experimentally
fragmented landscape. Ecology 76(5):1610–1624.
11. Cook WM, Lane KT, Foster BL, Holt RD (2002) Island theory, matrix effects and species
richness patterns in habitat fragments. Ecol Lett 5(5):619–623.
24. Bowman WD, Gartner JR, Holland K, Wiedermann M (2006) Nitrogen critical loads
for alpine vegetation and terrestrial ecosystem response: Are we there yet? Ecol
Appl 16(3):1183–1193.
25. Brewer S (2008) Declines in plant species richness and endemic plant species in
longleaf pine savannas invaded by Imperata cylindrica. Biol Invasions 10(8):
1257–1264.
26. Britton AJ, Beale CM, Towers W, Hewison RL (2009) Biodiversity gains and losses:
Evidence for homogenisation of Scottish alpine vegetation. Biol Conserv 142(8):
1728–1739.
27. Brunet J, Falkengren-Grerup U, Ruhling A, Tyler G (1997) Regional differences in
floristic change in South Swedish oak forests as related to soil chemistry and land
use. J Veg Sci 8(3):329–336.
28. Brunet J, Falkengren-Grerup U, Tyler G (1996) Herb layer vegetation of south
Swedish beech and oak forests - Effects of management and soil acidity during one
decade. For Ecol Manage 88(3):259–272.
29. Brunet J, Falkengren-Grerup U, Tyler G (1997) Pattern and dynamics of the ground
vegetation in south Swedish Carpinus betulus forests: Importance of soil chemistry
and management. Ecography 20(5):513–520.
30. Busing R (1993) Three decades of change at Albright Grove, Tennessee. Castanea
58(4):231–242.
31. Carey S, Ostling A, Harte J, del Moral R (2007) Impact of curve construction and
community dynamics on the species-time relationship. Ecology 88(9):2145–2153.
32. Clarke PJ, Latz PK, Albrecht DE (2005) Long-term changes in semi-arid vegetation:
Invasion of an exotic perennial grass has larger effects than rainfall variability. J Veg
Sci 16(2):237–248.
33. Clebsch E, Busing R (1989) Secondary succession, gap dynamics, and community
structure in a southern Appalachian cove forest. Ecology 70(3):728–735.
34. Collins S, Adams D (1983) Succession in grasslands: Thirty-two years of change in
a central Oklahoma tallgrass prairie. Plant Ecol 51(3):181–190.
35. Compagnoni A, Halpern CB (2009) Properties of native plant communities do not
determine exotic success during early forest succession. Ecography 32(3):449–458.
36. Corney PM, et al. (2008) Changes in the field-layer of Wytham Woods: Assessment of
the impacts of a range of environmental factors controlling change. J Veg Sci 19(3):
287–298.
37. Crow TR (1980) A rainforest chronicle: A 30-year record of change in structure and
composition at El Verde, Puerto Rico. Biotropica 12(1):42–55.
38. Damschen EI, Harrison S, Grace JB (2010) Climate change effects on an endemic-rich
edaphic flora: Resurveying Robert H. Whittaker’s Siskiyou sites (Oregon, USA).
Ecology 91(12):3609–3619.
39. Danby RK, Koh S, Hik DS, Price LW (2011) Four decades of plant community change
in the Alpine tundra of southwest Yukon, Canada. Ambio 40(6):660–671.
40. Daniels FJA, de Molenaar JG, Chytry M, Tichy L (2011) Vegetation change in
Southeast Greenland? Tasiilaq revisited after 40 years. Appl Veg Sci 14(2):230–241.
41. Davison SE, Forman RTT (1982) Herb and shrub dynamics in a mature oak forest: A
30-year study. Bull Torrey Bot Club 109(1):64–73.
42. de Villalobos AE, Zalba SM (2010) Continuous feral horse grazing and grazing
exclusion in mountain pampean grasslands in Argentina. Acta Oecol 36(5):514–519.
43. Debussche M, Debussche G, Lepart J (2001) Changes in the vegetation of Quercus
pubescens woodland after cessation of coppicing and grazing. J Veg Sci 12(1):81–92.
44. del Moral R (2007) Limits to convergence of vegetation during early primary
succession. J Veg Sci 18(4):479–488.
45. Doyle KM, Knight DH, Taylor DL, Barmore WJ, Benedict JM (1998) Seventeen years of
forest succession following the Waterfalls Canyon Fire in Grand Teton National Park,
Wyoming. Int J Wildland Fire 8(1):45–55.
46. Elliott KJ, Boring LR, Swank WT (1998) Changes in vegetation structure and diversity
after grass-to-forest succession in a Southern Appalachian watershed. Am Midl Nat
140(2):219–232.
47. Erschbamer B, Unterluggauer P, Winkler E, Mallaun M (2011) Changes in plant
species diversity revealed by long-term monitoring on mountain summits in the
Dolomites (northern Italy). Preslia 83(3):387–401.
2 of 6
48. Eschtruth AK, Cleavitt NL, Battles JJ, Evans RA, Fahey TJ (2006) Vegetation dynamics
in declining eastern hemlock stands: 9 years of forest response to hemlock woolly
adelgid infestation. Can J For Res 36(6):1435–1450.
49. Eskelinen A, Oksanen J (2006) Changes in the abundance, composition and species
richness of mountain vegetation in relation to summer grazing by reindeer. J Veg Sci
17(2):245–254.
50. Esselink P, Fresco LFM, Dijkema KS (2002) Vegetation change in a man-made salt
marsh affected by a reduction in both grazing and drainage. Appl Veg Sci 5(1):
17–32.
51. Falkengren-Grerup U (1986) Soil acidification and vegetation changes in deciduous
forest in southern Sweden. Oecologia 70(3):339–347.
52. Faust C, Storm C, Schwabe A (2012) Shifts in plant community structure of
a threatened sandy grassland over a 9-yr period under experimentally induced
nutrient regimes: Is there a lag phase? J Veg Sci 23(2):372–386.
53. Felfili JM, Rezende AV, Da Silva MC, Silva MA (2000) Changes in the floristic
composition of cerrado sensu stricto in Brazil over a nine-year period. J Trop Ecol
16(4):579–590.
54. Fischer M, Stocklin J (1997) Local extinctions of plants in remnants of extensively
used calcareous grasslands 1950-1985. Conserv Biol 11(3):727–737.
55. Fonty E, Sarthou C, Larpin D, Ponge J-F (2009) A 10-year decrease in plant species
richness on a neotropical inselberg: Detrimental effects of global warming? Glob
Change Biol 15(10):2360–2374.
56. Garbutt RA, Sparks TH (2002) Changes in the botanical diversity of a species rich
ancient hedgerow between two surveys (1971-1998). Biol Conserv 106(2):273–278.
57. Gibbens RP, Beck RF (1988) Changes in grass basal area and forb densities over a 64-
year period on grassland types of the Jornada Experimental Range. J Range Manage
41(3):186–192.
58. Grandin U (2011) Understorey vegetation stability and dynamics in unmanaged
boreal forests along a deposition gradient in Sweden. Ambio 40(8):867–877.
59. Gunnarsson U, Flodin L-A (2007) Vegetation shifts towards wetter site conditions on
oceanic ombrotrophic bogs in southwestern Sweden. J Veg Sci 18(4):595–604.
60. Gusewell S, Buttler A, Klotzli F (1998) Short-term and long-term effects of mowing
on the vegetation of two calcareous fens. J Veg Sci 9(6):861–872.
61. Hannus JJ, von Numers M (2010) Temporal changes in the island flora at different
scales in the archipelago of SW Finland. Appl Veg Sci 13(4):531–545.
62. Harrod J, White PS, Harmon ME (1998) Changes in xeric forests in western Great
Smoky Mountains National Park, 1936-1995. Castanea 63(3):346–360.
63. Hayasaka D, et al. (2012) Floristic variation of beach vegetation caused by the 2011
Tohoku-oki tsunami in northern Tohoku, Japan. Ecol Eng 44:227–232.
64. Heard MJ, Sax DF, Bruno JF (2012) Dominance of non-native species increases over
time in a historically invaded strandline community. Divers Distrib 18(12):1232–1242.
65. Hedl R (2004) Vegetation of beech forests in the Rychlebske Mountains, Czech
Republic, re-inspected after 60 years with assessment of environmental changes.
Plant Ecol 170(2):243–265.
66. Henderson RA, ed (1992) Ten-Year Response of a Wisconsin Prairie Remnant to
Seasonal Timing of Fire (Univ of Northern Iowa, Cedar Falls, IA), Vol 12, pp 121-125.
67. Holland PG (1978) Species turnover in deciduous forest vegetation. Vegetatio 38(2):
113–118.
68. Hollister RD, Webber PJ, Tweedie CE (2005) The response of Alaskan arctic tundra to
experimental warming: Differences between short- and long-term responses. Glob
Change Biol 11(4):525–536.
69. Huberty LE, Gross KL, Miller CJ (1998) Effects of nitrogen addition on successional
dynamics and species diversity in Michigan old-fields. J Ecol 86(5):794–803.
70. Hudson JMG, Henry GHR (2010) High Arctic plant community resists 15 years of
experimental warming. J Ecol 98(5):1035–1041.
71. Jerling L (1998) Linnaeus’ flora Kofsoensis revisited: Floristic changes during 260
years in a small island of the Lake Maelaren. Nord J Bot 18(6):667–680.
72. Jiménez MA, et al. (2011) Extreme climatic events change the dynamics and
invasibility of semi-arid annual plant communities. Ecol Lett 14(12):1227–1235.
73. Kapfer J, Grytnes JA, Gunnarsson U, Birks HJB (2011) Fine-scale changes in
vegetation composition in a boreal mire over 50 years. J Ecol 99(5):1179–1189.
74. Kiehl K, Schröder H, Stock M (2007) Long-term vegetation dynamics after land-use
change in Wadden Sea salt marshes. Coastline Reports 7:17–24.
75. Korb JE, Ranker TA (2001) Changes in stand composition and structure between
1981 and 1996 in four Front Range plant communities in Colorado. Plant Ecol 157(1):
1–11.
76. Kralovec J, Pocova L, Jonasova M, Macek P, Prach K (2009) Spontaneous recovery of
an intensively used grassland after cessation of fertilizing. Appl Veg Sci 12(4):
391–397.
77. Kraszewski SE, Waller DM (2008) Fifty-five year changes in species composition on
dry prairie remnants in south-central Wisconsin. J Torrey Bot Soc 135(2):236–244.
78. Kuiters AT, Kramer K, Van der Hagen H, Schaminee JHJ (2009) Plant diversity, species
turnover and shifts in functional traits in coastal dune vegetation: Results from
permanent plots over a 52-year period. J Veg Sci 20(6):1053–1063.
79. Lameire S, Hermy M, Honnay O (2000) Two decades of change in the ground
vegetation of a mixed deciduous forest in an agricultural landscape. J Veg Sci 11(5):
695–704.
80. Laughlin DC, Moore MM, Fulé PZ (2011) A century of increasing pine density and
associated shifts in understory plant strategies. Ecology 92(3):556–561.
81. Leendertse PC, Roozen AJM, Rozema J (1997) Long-term changes (1953-1990) in the
salt marsh vegetation at the Boschplaat on Terschelling in relation to sedimentation
and flooding. Plant Ecol 132(1):49–58.
82. Lima ED, Lima HS, Ratter JA (2009) Post fire changes in the structure and
composition of the woody vegetation of a mesotrophic cerrado over a five-year
period (1997-2002) in Nova Xavantina, MT, Brazil. Cerne 15(4):468–480.
Vellend et al. www.pnas.org/cgi/content/short/1312779110
83. Lu X, Mo J, Gilliam FS, Zhou G, Fang Y (2010) Effects of experimental nitrogen
additions on plant diversity in an old-growth tropical forest. Glob Change Biol
16(10):2688–2700.
84. Lu XK, et al. (2011) Effects of experimental nitrogen additions on plant diversity in
tropical forests of contrasting disturbance regimes in southern China. Environ Pollut
159(10):2228–2235.
85. Matus G, Papp M, Tothmeresz B (2005) Impact of management on vegetation
dynamics and seed bank formation of inland dune grassland in Hungary. Flora
200(3):296–306.
86. McGovern S, Evans CD, Dennis P, Walmsley C, McDonald MA (2011) Identifying
drivers of species compositional change in a semi-natural upland grassland over a
40-year period. J Veg Sci 22(2):346–356.
87. Meiners SJ, Pickett STA, Cadenasso ML (2002) Exotic plant invasions over 40 years of
old field successions: Community patterns and associations. Ecography 25(2):
215–223.
88. Mews H, Marimon B, Ratter J (2012) Observations on the vegetation of Mato Grosso,
Brazil. V. Changes in the woody species diversity of a forest in the Cerrado–
Amazonian forest transition zone and notes on the forests of the region. Edinb J Bot
69(2):239–253.
89. Milbauer ML, Leach MK (2007) Influence of species pool, fire history, and woody
canopy on plant species density and composition in tallgrass prairie. J Torrey Bot Soc
134(1):53–62.
90. Moles R, Breen J, O’Regan B (2005) A pilot scale long-term experimental study on the
effects of grazing and gap creation on burren grassland dynamics: Implications for
conservation. Biol Environ Proc R Irish Acad 105(1):15-32.
91. Morrison LW (2010) Long-term non-equilibrium dynamics of insular floras: A 17-year
record. Glob Ecol Biogeogr 19(5):663–672.
92. Myers BR, Walck JL, Blum KE (2004) Vegetation change in a former chestnut stand
on the Cumberland Plateau of Tennessee during an 80-year period (1921-2000).
Castanea 69(2):81–91.
93. Myster RW, Malahy MP (2008) Is there a middle way between permanent plots and
chronosequences? Can J For Res 38(12):3133–3138.
94. Naaf T, Wulf M (2010) Habitat specialists and generalists drive homogenization and
differentiation of temperate forest plant communities at the regional scale. Biol
Conserv 143(4):848–855.
95. Nygaard PH, Odegaard T (1999) Sixty years of vegetation dynamics in a south boreal
coniferous forest in southern Norway. J Veg Sci 10(1):5–16.
96. Odland A, del Moral R (2002) Thirteen years of wetland vegetation succession
following a permanent drawdown, Myrkdalen Lake, Norway. Plant Ecol 162(2):
185–198.
97. Okland RH, Eilertsen O (1996) Dynamics of understory vegetation in an old-growth
boreal coniferous forest, 1988-1993. J Veg Sci 7(5):747–762.
98. Palmer MW, McAlister SD, Arevalo JR, DeCoster JK (2000) Changes in the understory
during 14 years following catastrophic windthrow in two Minnesota forests. J Veg Sci
11(6):841–854.
99. Panitsa M, Tzanoudakis D, Sfenthourakis S (2008) Turnover of plants on small islets of
the eastern Aegean Sea within two decades. J Biogeogr 35(6):1049–1061.
100. Pascarella JB, Aide TM, Zimmerman JK (2004) Short-term response of secondary
forests to hurricane disturbance in Puerto Rico, USA. For Ecol Manage 199(2-3):
379–393.
101. Pauli H, Gottfried M, Reiter K, Klettner C, Grabherr G (2007) Signals of range
expansions and contractions of vascular plants in the high Alps: Observations (1994-
2004) at the GLORIA*master site Schrankogel, Tyrol, Austria. Glob Change Biol 13(1):
147–156.
102. Pellerin S, Mercure M, Desaulniers AS, Lavoie C (2009) Changes in plant communities
over three decades on two disturbed bogs in southeastern Quebec. Appl Veg Sci
12(1):107–118.
103. Perry JE, Hershner CH (1999) Temporal changes in the vegetation pattern in a tidal
freshwater marsh. Wetlands 19(1):90–99.
104. Peter M, Edwards PJ, Jeanneret P, Kampmann D, Luscher A (2008) Changes over
three decades in the floristic composition of fertile permanent grasslands in the
Swiss Alps. Agric Ecosyst Environ 125(1-4):204–212.
105. Peter M, Gigon A, Edwards PJ, Luscher A (2009) Changes over three decades in the
floristic composition of nutrient-poor grasslands in the Swiss Alps. Biodivers Conserv
18(3):547–567.
106. Pettit NE, Froend RH (2001) Long-term changes in the vegetation after the cessation
of livestock grazing in Eucalyptus marginata (jarrah) woodland remnants. Austral
Ecol 26(1):22–31.
107. Piessens K, Aerts N, Hermy M (2006) Long-term (1978-2003) effects of an extensive
grazing regime on plant species composition of a heathland reserve. Belg J Bot
139(1):49–64.
108. Pysek P, et al. (2004) Trends in species diversity and composition of urban vegetation
over three decades. J Veg Sci 15(6):781–788.
109. Rogers DA, Rooney TP, Olson D, Waller DM (2008) Shifts in southern Wisconsin forest
canopy and understory richness, composition, and heterogeneity. Ecology 89(9):
2482–2492.
110. Roitman I, Felfili JM, Rezende AV (2008) Tree dynamics of a fire-protected cerrado
sensu stricto surrounded by forest plantations, over a 13-year period (1991-2004) in
Bahia, Brazil. Plant Ecol 197(2):255–267.
111. Rooney TP, Wiegmann SM, Rogers DA, Waller DM (2004) Biotic impoverishment
and homogenization in unfragmented forest understory communities. Conserv Biol
18(3):787–798.
112. Runkle JR, DiSalvo A, Graham-Gibson Y, Dorning M (2007) Vegetation release eight
years after removal of Lonicera maackii in West-Central Ohio. Ohio J Sci 107(5):
125–129.
3 of 6
113. Russell-Smith J, Price OF, Murphy BP (2010) Managing the matrix: Decadal responses
of eucalypt-dominated savanna to ambient fire regimes. Ecol Appl 20(6):1615–1632.
114. Saiter FZ, Guilherme FAG, Thomaz LD, Wendt T (2011) Tree changes in a mature
rainforest with high diversity and endemism on the Brazilian coast. Biodivers
Conserv 20(9):1921–1949.
115. Sang W, Bai F (2009) Vascular diversity patterns of forest ecosystem before and after
a 43-year interval under changing climate conditions in the Changbaishan Nature
Reserve, northeastern China. Plant Ecol 201(1):115–130.
116. Schütz M, Risch AC, Leuzinger E, Krüsi BO, Achermann G (2003) Impact of herbivory
by red deer (Cervus elaphus L.) on patterns and processes in subalpine grasslands in
the Swiss National Park. For Ecol Manage 181(1-2):177–188.
117. Shaver GR, et al. (2001) Species composition interacts with fertilizer to control long-
term change in tundra productivity. Ecology 82(11):3163–3181.
118. Smart SM, et al. (2006) Biotic homogenization and changes in species diversity across
human-modified ecosystems. Proc Biol Sci 273(1601):2659–2665.
119. Snogerup S, Snogerup B (2004) Changes in the flora of some Aegean islets 1968-
2000. Plant Syst Evol 245(3-4):169–213.
120. Song L, et al. (2011) Nitrogen enrichment enhances the dominance of grasses over
forbs in a temperate steppe ecosystem. Biogeosciences 8(8):2341–2350.
121. Specht R, Connor D, Clifford H (1977) The heath-savannah problem: The effect of
fertilizer on sand-heath vegetation of North Stradbroke Island, Queensland. Aust J
Ecol 2(2):179–186.
122. Sundaram B, Hiremath AJ (2012) Lantana camara invasion in a heterogeneous
landscape: Patterns of spread and correlation with changes in native vegetation.
Biol Invasions 14(6):1127–1141.
123. Taverna K, Peet RK, Phillips LC (2005) Long-term change in ground-layer vegetation
of deciduous forests of the North Carolina Piedmont, USA. J Ecol 93(1):202–213.
124. Taylor D, Hamilton AC, Lewis SL, Nantale G (2008) Thirty-eight years of change in
a tropical forest: Plot data from Mpanga Forest Reserve, Uganda. Afr J Ecol 46(4):
655–667.
125. Thiemann JA, et al. (2009) Herbaceous-layer impoverishment in a post-agricultural
southern Appalachian landscape. Am Midl Nat 162(1):148–168.
126. Thimonier A, Dupouey JL, Bost F, Becker M (1994) Simultaneous eutrophication and
acidification of a forest ecosystem in North-East France. New Phytol 126(3):533–539.
127. Thimonier A, Dupouey JL, Timbal J (1992) Floristic changes in the herb-layer
vegetation of a deciduous forest in the Lorraine Plain under the influence of
atmospheric deposition. For Ecol Manage 55(1-4):149–167.
128. Thomaz SM, Carvalho P, Padial AA, Kobayashi JT (2009) Temporal and spatial
patterns of aquatic macrophyte diversity in the Upper Paraná River floodplain. Braz
J Biol 69(2, Suppl):617–625.
129. Todd SW, Hoffman MT (2009) A fence line in time demonstrates grazing-induced
vegetation shifts and dynamics in the semiarid Succulent Karoo. Ecol Appl 19(7):
1897–1908.
130. Trabaud L, Lepart J (1981) Changes in the floristic composition of a Quercus coccifera
L. garrigue in relation to different fire regimes. Vegetatio 46-47(1):105–116.
Species richness (155)
exp(Shannon) (35)
Evenness (18)
-0.2
Effect size (log(metricY2/metricY1).decade-1)
Fig. S1. Comparison of mean effect sizes ±95% credible intervals or 95% bias-corrected CIs in the multilevel analysis and the conventional meta-analysis,
respectively. Results are shown for species richness (multilevel analysis shown in the main text), and for the “numbers equivalent” of the Shannon index of
diversity [exp(Shannon)] and Pielou’s index of evenness. Bias-corrected CIs are generated via bootstrapping and thus may be asymmetric. See Table S2 and SI
Methods for details of analysis.
Vellend et al. www.pnas.org/cgi/content/short/1312779110
131. Tsuyuzaki S (2009) Causes of plant community divergence in the early stages of
volcanic succession. J Veg Sci 20(5):959–969.
132. Vacek S, Bastl M, Leps J (1999) Vegetation changes in forests of the Krkonose Mts.
over a period of air pollution stress (1980-1995). Plant Ecol 143(1):1–11.
133. Van Calster H, et al. (2007) Management driven changes (1967-2005) in soil acidity
and the understorey plant community following conversion of a coppice-with-
standards forest. For Ecol Manage 241(1-3):258–271.
134. Van Calster H, et al. (2008) Diverging effects of overstorey conversion scenarios on
the understorey vegetation in a former coppice-with-standards forest. For Ecol
Manage 256(4):519–528.
135. van den Berg LJL, et al. (2011) Direct and indirect effects of nitrogen deposition on
species composition change in calcareous grasslands. Glob Change Biol 17(5):1871–1883.
136. Vandermeer J, Granzow de la Cerda I, Boucher D, Perfecto I, Ruiz J (2000) Hurricane
disturbance and tropical tree species diversity. Science 290(5492):788–791.
137. Vennetier M, Ripert C (2009) Forest flora turnover with climate change in the
Mediterranean region: A case study in Southeastern France. For Ecol Manage 258:
S56–S63.
138. Verburg R, van Eijk-Bos C (2003) Effects of selective logging on tree diversity,
composition and plant functional type patterns in a Bornean rain forest. J Veg Sci
14(1):99–110.
139. Verheyen K, et al. (2012) Driving factors behind the eutrophication signal in
understorey plant communities of deciduous temperate forests. J Ecol 100(2):
352–365.
140. Villarreal S, et al. (2012) Tundra vegetation change near Barrow, Alaska (1972-2010).
Environ Res Lett 7(1):15508.
141. Vittoz P, Randin C, Dutoit A, Bonnet F, Hegg O (2009) Low impact of climate change
on subalpine grasslands in the Swiss Northern Alps. Glob Change Biol 15(1):209–220.
142. von Oheimb G, Brunet J (2007) Dalby Soderskog revisited: Long-term vegetation
changes in a south Swedish deciduous forest. Acta Oecologica-International Journal
of Ecology 31(2):229–242.
143. Wesche K, Krause B, Culmsee H, Leuschner C (2012) Fifty years of change in Central
European grassland vegetation: Large losses in species richness and animal-
pollinated plants. Biol Conserv 150(1):76–85.
144. Wesche K, Ronnenberg K, Retzer V, Miehe G (2010) Effects of large herbivore exclusion
on southern Mongolian desert steppes. Acta Oecolog Int J Ecol 36(2):234–241.
145. Wilson SD, Nilsson C (2009) Arctic alpine vegetation change over 20 years. Glob
Change Biol 15(7):1676–1684.
146. Wiser SK, Allen RB, Clinton PW, Platt KH (1998) Community structure and forest
invasion by an exotic herb over 23 years. Ecology 79(6):2071–2081.
147. Woods KD, Hicks DJ, Schultz J (2012) Losses in understory diversity over three
decades in an old-growth cool-temperate forest in Michigan, USA. Can J For Res
42(3):532–549.
148. Zhao DH, Allen B, Sharitz RR (2006) Twelve year response of old-growth
southeastern bottomland hardwood forests to disturbance from Hurricane Hugo.
Can J For Res 36(12):3136–3147.
Approach
Multilevel
Conventional meta-analysis
0.0 0.2
4 of 6
 Number of studies
15

10

Raw effect size (log(SRY2/SRY1))

1.0

0.5

0.0

-0.5

-1.0

10 100 0 10 20
Study duration (years) Number of studies

Fig. S2. Relationship between study duration (log scale) and raw species richness effect sizes. Histograms of the two variables are shown on the outer edges of
each axis. The histogram of raw effect sizes (right edge) reveals the same result of richness changes centered on zero as we found using the effect sizes
calculated to express change per decade (Fig. 2).

Table S1. Summary of the BUGS output of the multilevel models of study-level effect sizes
μES Quantiles μES (%)

Data subset N Mean SE τ 2.5 12.5 25 75 87.5 97.5

All data 155 (318) 0.032 0.024 0.258 −0.018 0.005 0.016 0.048 0.060 0.081
Habitat
Forest 74 (157) 0.021 0.035 0.271 −0.047 −0.019 −0.003 0.044 0.061 0.092
Grassland 39 (75) 0.014 0.051 0.261 −0.085 −0.046 −0.019 0.049 0.072 0.114
Savanna/shrubland 14 (25) 0.037 0.053 0.179 −0.070 −0.022 0.004 0.069 0.094 0.145
Tundra 7 (21) −0.038 0.071 0.137 −0.185 −0.108 −0.076 0.003 0.032 0.104
Wetland 15 (29) −0.002 0.103 0.236 −0.199 −0.114 −0.071 0.062 0.115 0.214
Continent
Europe 78 (179) −0.003 0.019 0.104 −0.041 −0.025 −0.016 0.011 0.019 0.036
North America 49 (87) 0.039 0.059 0.374 −0.076 −0.027 0.0004 0.078 0.106 0.160
South America 12 (18) 0.192 0.137 0.307 −0.082 0.046 0.105 0.279 0.340 0.454
Asia 9 (22) 0.063 0.179 0.481 −0.295 −0.129 −0.043 0.169 0.249 0.414
Australia 5 (9) 0.091 0.190 0.279 −0.236 −0.054 0.013 0.165 0.243 0.435
Africa 2 (3) Raw mean = 0.165
Driver
Postdisturbance 13 (23) 0.355 0.147 0.365 0.054 0.195 0.265 0.446 0.516 0.644
Postfire 8 (12) 0.158 0.124 0.211 −0.089 0.030 0.087 0.230 0.287 0.404
Ongoing disturbance 6 (10) 0.030 0.115 0.141 −0.206 −0.087 −0.034 0.095 0.145 0.261
Ongoing fire 6 (19) −0.014 0.115 0.240 −0.248 −0.129 −0.078 0.051 0.101 0.209
Herbivory/grazing 12 (17) 0.040 0.039 0.089 −0.040 −0.003 0.015 0.066 0.083 0.115
Cessation of grazing 7 (45) 0.101 0.162 0.374 −0.227 −0.064 0.012 0.190 0.260 0.405
Canopy closure 17 (18) 0.017 0.025 0.078 −0.034 −0.012 0.001 0.033 0.045 0.067
Climate change 11 (33) −0.063 0.049 0.132 −0.159 −0.118 −0.094 −0.033 −0.010 0.037
Pollution 8 (11) −0.025 0.054 0.099 −0.129 −0.082 −0.058 0.006 0.034 0.082
Invasion 7 (10) −0.357 0.323 0.739 −0.991 −0.687 −0.549 −0.167 −0.028 0.263
Management 16 (41) −0.008 0.053 0.109 −0.113 −0.067 −0.043 0.026 0.053 0.098
Unknown 45 (71) 0.021 0.045 0.246 −0.068 −0.031 −0.010 0.052 0.073 0.109

The mean effect sizes (μES) express the grand mean decadal-scale species richness log ratios, as described in
the main text, and the SE expresses the uncertainty about this mean. τ is the superpopulation SD, expressing the
heterogeneity among studies. N is the number of studies (data sets) in the analysis. Bold numbers indicate means
with 95% credible intervals that do not overlap zero.

Vellend et al. www.pnas.org/cgi/content/short/1312779110 5 of 6

 Table S2. Results of the conventional fixed-effect
meta-analyses
Diversity index N Weighted mean 95% bias-corrected CI

Species richness 155 0.017 −0.023 0.053

exp(Shannon) 35 −0.007 −0.084 0.130
Evenness 18 −0.050 −0.175 0.020

In this analysis we estimated the weighted mean of the log ratio (effect
size) under the assumption of a single effect size common to all studies (1, 2).
As most studies had several data lines in the database, species richness, di-
versity, and evenness values were pooled at the study level before this anal-
ysis by calculating the mean, weighted by sample size. N is the number of
studies in the analysis. Because calculating the variance of the log ratio
requires the covariance between the nonindependent SRY1,i and SRY2,i values
(3), which was never reported in the original studies, we used the square
root of the sample size as a weighting factor for the resampling test (4). The
bias corrected 95% CIs were calculated with 2,000 bootstrap resamples (1, 4).

1. Hedges LV, Gurevitch J, Curtis PS (1999) The meta-analysis of response ratios in experimental ecology. Ecology 80(4):1150–1156.
2. Harrison F (2011) Getting started with meta-analysis. Meth Ecol Evol 2(1):1–10.
3. Lajeunessei MJ (2011) On the meta-analysis of response ratios for studies with correlated and multi-group designs. Ecology 92(11):2049–2055.
4. Adams DC, Gurevitch J, Rosenberg MS (1997) Resampling tests for meta-analysis of ecological data. Ecology 78(4):1277–1283.

Table S3. Results of multilevel analysis with covariates

Model coefficient (γ 1)

Predictor variable Mean or no. studies Median Range Mean 95% CI

Latitude (°) 39.3 46.0 −43.2, 78.9 −0.0016 −0.0036 0.0005

Longitude (°) −19.2 4.1 −156.7, 172.8 −0.0001 −0.0008 0.0006
Plot area (m2) 653.5 43.8 0.04, 13,000 0.0004 −0.0019 0.0026
Time interval (y) 25.9 18.0 5, 261 −0.0019 −0.0042 0.0005
Plot permanent (no/yes) 59/96 0.0715 −0.0249 0.1683

Effect sizes across the j studies were modeled as ESj ∼ N(γ 0 + γ 1Uj, τ2), in which Uj is a study-level covariates (1).
Iterations were set as for the main analysis and priors of γ 0 and γ 1 ∼ N(0, 10,000). ESj was expressed per decade
(main text) for all analyses except when using Time interval as a predictor, in which case the raw log ratio was
used. Plot area was square root transformed before analysis. The reference level for “Plot permanent” was set as
no; the coefficient for this analysis models the increase in effect size in permanent plots. Note that the means
here were calculated across studies, such that some means reported across data sets in the main text are slightly
different (some studies included multiple data sets).

1. Gelman A, Hill J (2007) Data Analysis Using Regression and Multilevel/Hierarchical Models (Cambridge Univ Press, Cambridge, MA).

Other Supporting Information Files

Dataset S1 (XLS)

Vellend et al. www.pnas.org/cgi/content/short/1312779110 6 of 6